Molecular and circuitry mechanism underlying autism behaviors in Shank3 mouse models ABSTRACT: While there has been significant progress in identifying the genetic defects in individuals with autism spectrum disorders (ASD), the molecular heterogeneity fails to provide mechanistic insights into how genetic defects result in ASD. It has been proposed that defects in >100 genes strongly implicated in ASD converge onto a few distinct molecular pathways and contribute to the shared neural circuits supporting ASD. To identify the underlying mechaniams we adopted the RDoC matrix to establish links within a Gene-Molecule-Circuit-Behavior (GMCB) axis for SHANK3-related ASD. Defects in SHANK3 are found in ~2% of individuals with ASD, and the most common defect (>95%) is deletion of the entire SHANK3 gene. In contrast to the 14 Shank3 isoform-specific knockout (KO) mice, we have generated a unique Shank3 complete-KO mouse by conventional or conditional deletion of exons 4-22 (?e4-22 or e4-22flox). The ?e4-22 mice display robust behaviors that recapitulate core features of ASD associated with SHANK3-deficiency in humans. We recently generated a new mouse with a Homer1 binding mutation (SH3-PL) in Shank3 that permits us to probe mechanistic links within the RDoC matrix. In ?e4-22 mice functions of cortical/striatal synapses are impaired and functional connectivity is abnormal in the nucleus acumens (NAC) -associated axis. Results from striatal- and cortical-specific Shank3 e4-22flox mice and viral-mediated SHANK3 rescue suggest distinct roles for the NAC and cortical circuits in social, instrumental, and repetitive behaviors. Molecularly, Homer1-mGluR5 scaffolds are altered in the striatum but not cortex, while NMDA receptors are reduced in cortex but not striata of ?e4-22 mice. Our central hypothesis is that abnormal social and repetitive behaviors in Shank3 mice are caused by alterations in Homer1-mGluR5 scaffolds in a NAC circuit and in NMDAR functions in a cortical circuit, respectively. These molecular- and circuit-specific aberrations may represent a convergent molecular pathway with shared neural circuits that underlie the abnormalities in social, instrumental, and repetitive behaviors. The specific objective of this proposal is to use the RDoC matrix as guide to analyze mechanisms within a GMCB axis using our Shank3 complete-KO and the new SH3PL mice. RELEVANCE: These results will provide unique insights into the roles that Homer1-mGluR5 scaffolds and NMDARs play in ASD-like behaviors by establishing a convergent point of genes, molecular pathways, and neural circuits that may be amenable to targeted treatments for ASD.